Therapeutic drug monitoring for optimizing amisulpride therapy in patients with Schizophrenia

Amisulpride is a clinically effective antipsychotic drug in a broad dose range with low propensity for extrapyramidal symptoms (EPS). Daily doses and plasma levels of amisulpride were analyzed within a large-scale therapeutic drug monitoring (TDM) survey to find plasma level ranges for optimized treatment under naturalistic conditions. Data of 378 schizophrenic patients treated with amisulpride (100–1550 mg) were included (40% female). Amisulpride plasma levels were analyzed at steady state; assessment comprised improvement (CGI-I) and side-effects, particularly EPS. For detection of cut-off values regarding non-response or EPS, receiver operating characteristics (ROC) curves were applied and the area under the ROC curve (AUC) was calculated. Amisulpride daily doses (594 ± 262 mg) and plasma levels (315 ± 277 ng/ml) were significantly correlated (r = 0.53; P < 0.0001). Patients with non-response to amisulpride (8.9%) had significantly (P < 0.05) lower plasma levels (248 ± 291 ng/ml) than patients with at least moderate improvement (316 ± 253 ng/ml) despite comparable amisulpride doses (628 ± 253 vs. 590 ± 263 mg). Patients with EPS (14.6%) had significantly (P < 0.05) higher amisulpride plasma levels (377 ± 290 ng/ml) than patients without EPS (305 ± 274 ng/ml) despite similar doses in both groups (595 ± 266 vs. 594 ± 246 mg). ROC analyses revealed significant predictive properties of amisulpride plasma levels (P < 0.05) for non-response (AUC = 0.65 ± 0.05) and EPS (AUC = 0.62 ± 0.05), respectively. Daily amisulpride doses did not significantly predict non-response or EPS. Optimal amisulpride plasma level values to avoid non-response and EPS were ⩾100 or ⩽320 ng/ml, respectively. Analysis of clinical utility revealed that blood levels must be analyzed in 7 patients until one patient benefits from the TDM procedure by avoiding non-response or EPS. Although our results were mainly explorative, TDM of amisulpride seems very useful for clinical decision making.